Search found 20 matches

by Robert McGibbon
Thu Oct 29, 2015 3:16 pm
Forum: OpenMM
Topic: Not using checkpoint files correctly(?)
Replies: 16
Views: 2793

Re: Not using checkpoint files correctly(?)

Did you compile OpenMM from source? If so, what version of `swig` did you use?
by Robert McGibbon
Mon Jun 10, 2013 4:03 pm
Forum: OpenMM
Topic: OpenMM5.1 'first periodic box must be parallel to x' problem
Replies: 6
Views: 2302

Re: OpenMM5.1 'first periodic box must be parallel to x' pro

I've updated the script builder (see https://github.com/rmcgibbo/openmm-webbuilder/issues/8) to issue a warning in this scenario.
by Robert McGibbon
Mon Apr 08, 2013 12:59 pm
Forum: OpenMM
Topic: bonded forces with variable numbers of particles
Replies: 9
Views: 1341

Re: bonded forces with variable numbers of particles

Just to throw this out there, I agree w/ Justin that it would be really nice if it were possible for all of the plugin code, including stuff for the python wrappers and CMAKE stuff, to be containing within a single directory. I'm not sure how feasible this is for a CMake though.

-Robert
by Robert McGibbon
Wed Feb 27, 2013 3:46 pm
Forum: MSMBuilder
Topic: Sign up for msmbuilder-user email list
Replies: 0
Views: 791

Sign up for msmbuilder-user email list

We've set up an email list for users of msmbuilder. If you use msmbuilder, please sign up. This will be a convenient place for people inside and outside of the Pande group to ask and answer questions about using msmbuilder and building models. You can sign up with this link: https://mailman.stanford...
by Robert McGibbon
Wed Feb 27, 2013 3:19 pm
Forum: MSMBuilder
Topic: PCCA+ eigenvalue convergence issue
Replies: 1
Views: 1117

Re: PCCA+ eigenvalue convergence issue

Ali, My first reaction is that a lag time of 1ps seems much to short for a large protein like barnase. We're in a very different regime here -- 110 resides -- than we are with the alanine dipeptide tutorial. Did you subsample further? Did both PCCA and PCCA+ have the eigenvalue convergence issue? Al...
by Robert McGibbon
Wed Feb 27, 2013 3:14 pm
Forum: MSMBuilder
Topic: Problems with CalculateImpliedTimescales.py
Replies: 1
Views: 1161

Re: Problems with CalculateImpliedTimescales.py

Shan, That error message means that the model produced only has two states, and as such cannot have 10 timescales -- each timescale corresponds to an eigenvector of the transition matrix and a 2x2 matrix can only have 2 linearly independent eigenvectors. This is probably being caused by the ergodic ...
by Robert McGibbon
Wed Feb 27, 2013 3:10 pm
Forum: MSMBuilder
Topic: fail to compute implied timescale after hybrid clustering
Replies: 1
Views: 1083

Re: fail to compute implied timescale after hybrid clusterin

Lizhe, This looks like a shape error. Perhaps your assignments array 1 dimensional instead of the required 2 dimensional? We (the developers) have been trying to improve the quality of the error messages in the 2.6 release. Also, sorry for not checking this forum for a very long time. Please sign up...
by Robert McGibbon
Wed Feb 27, 2013 3:06 pm
Forum: MSMBuilder
Topic: why the plot implied time scale vs lag time is very bumpy?
Replies: 2
Views: 1844

Re: why the plot implied time scale vs lag time is very bump

Junyan, This is probably just statistical noise. The more clusters you have, the fewer number of transitions there are per pair of states, the the higher variance we should expect any statistical estimators to be. Also, sorry for not checking this forum for a very long time. Please sign up for the m...
by Robert McGibbon
Tue Oct 02, 2012 12:19 pm
Forum: MSMBuilder
Topic: Contribute to MSMBuilder Development on Github!
Replies: 0
Views: 909

Contribute to MSMBuilder Development on Github!

The MSMBuilder development team is now using GitHub to host code and coordinate development activity. The repository is publicly accessible at https://github.com/SimTk/msmbuilder That site is now the main venues through which the developers are communicating and planning new features (and fixing bug...
by Robert McGibbon
Tue Oct 02, 2012 12:11 pm
Forum: MSMBuilder
Topic: pair wise RMSD calculation
Replies: 1
Views: 1040

Re: pair wise RMSD calculation

Thanks for the bug report! No, you're not missing anything -- this is just a bug on our end. If you remove that line in the source code for the script (in scripts/CalculatePairwiseRMSD.py) and then reinstall the package, it should work. As you can see from grepping the source, the HierarchicalCluste...